Abundance of selected fish species in relation to temperature and salinity patterns in ... Alaska, following construction of the Endicott Causeway.
2 LGL Ecological Research Associates, Inc., 1410 Cavitt Street, Bryan, Texas 77801, U.S.A.
3 LGL Alaska, Research Associates, Inc., 4175 Tudor Centre Drive, Suite 101, Anchorage, Alaska 99508, U.S.A.
(Received 9 July 1997; accepted in revised form 23 October 1997)
ABSTRACT. Hydrography data from 1985 to 1993 in the vicinity of the Endicott Causeway near Prudhoe Bay, Alaska, were analyzed to determine if the trend in yearly post-causeway construction temperature and salinity was toward colder and more saline conditions. No significant (p <= 0.05) relationship was found for temperature; however, salinity levels significantly decreased with time following construction. Intra-year variations in hydrographic conditions appeared to be due to coast-wide meteorological and oceanographic events rather than causeway-induced. We analyzed abundance indices, as logetransformed catch-per-unit-effort (CPUE), of four marine species (arctic cod, Boreogadus saida; fourhorn sculpin, Myoxocephalus quadricornis; arctic flounder, Pleuronectes glacialis; saffron cod, Eleginus gracilis); two freshwater species (round whitefish, Prosopium cylindraceum; arctic grayling, Thymallus arcticus); and the anadromous rainbow smelt, Osmerus mordax, to determine whether there were increases in the abundances of marine species and declines in anadromous and freshwater species. We developed the "best-fit" regression models, based on the criterion of maximized regression F-ratio (i.e., minimized regression p-values), using year, temperature, salinity, and all possible interactions. Results suggested that all species exhibited changes in abundance levels over the period of record. In particular, arctic flounder and rainbow smelt both showed sharp increases in abundance five years after construction. For all species, the observed changes in abundance levels appeared to be due to naturally occurring events in the species' life histories or changes in coast-wide meteorological conditions rather than to causeway-induced alterations in local hydrographic conditions.
Key words: Alaska, Sagavanirktok Delta, causeway, arctic fishes, abundance, temperature, salinity
RESUME. On a analyse les donnees hydrographiques obtenues de 1985 a 1993 dans le voisinage immediat du pont-jetee Endicott situe pres de Prudhoe Bay en Alaska, en vue de determiner si, apres la construction du pont, la temperature annuelle avait tendance a etre plus froide et l'eau plus saline. On n'a trouve aucun rapport significatif (p <= 0,05) pour la temperature. Le niveau de salinite, par contre, a considerablement baisse a mesure que le temps passait apres la construction. Les variations intra-annuelles dans les conditions hydrographiques semblaient etre dues a des phenomenes meteorologiques et oceanographiques affectant le littoral tout entier plutot qu'au pont-jetee lui-meme. On a analyse les indices d'abondance, exprimes en loge des prises par unite d'effort (PPUE), de quatre especes marines (morue arctique, Boreogadus saida; chaboisseau a quatre cornes, Myoxocephalus quadricornis; plie arctique, Pleuronectes glacialis; navaga jaune, Eleginus gracilis); deux especes d'eau douce (menomini rond, Prosopium cylindraceum, et ombre arctique, Thymallus arcticus); ainsi qu'une espece anadrome (eperlan arc-en-ciel, Osmerus mordax), afin de determiner s'il y avait une augmentation dans l'abondance des especes marines et une diminution dans les especes anadromes et d'eau douce. On a developpe des modeles de regression de meilleur ajustement, fondes sur le critere de maximisation du F de regression (c.-a-d. minimum des valeurs p de regression), en utilisant l'annee, la temperature, la salinite et toutes les interactions possibles. Les resultats suggeraient que toutes les especes montraient des changements dans les niveaux d'abondance durant la periode de releves. La plie arctique et l'eperlan arc-en-ciel en particulier affichaient tous deux un accroissement tres net dans leur abondance cinq ans apres la construction. Pour toutes les especes, les changements observes dans les niveaux d'abondance semblent etre dus a des phenomenes qui se produisent naturellement dans le cycle biologique de l'espece ou a des changements dans les conditions meteorologiques affectant le littoral tout entier plutot qu'a des transformations causees par le pont-jetee dans les conditions hydrographiques locales.
Mots cles: Alaska, delta du Sagavanirktok, pont-jetee, poissons de l'Arctique, abondance, temperature, salinite
Traduit pour la revue Arctic par Nesida Loyer.
For over 15 years, fish and hydrographic studies have been conducted in the Prudhoe Bay region, Alaska, to monitor the effects of nearshore oil and gas developments on fish populations. The studies have resulted in several publications about the distribution, movement patterns, and growth rates of various fish species (Craig et al., 1982, 1985; Craig, 1984, 1989; Fechhelm et al., 1989, 1992, 1995; Griffiths et al., 1992). In addition, there have been studies of the effects of nearshore developments on the local hydrography and fish populations (Niedoroda and Colonell, 1990; Cannon et al., 1991; Gallaway et al., 1991; Hachmeister et al., 1991), and methods have been developed for synthesizing available data to address specific fish-related development issues (Wilson and Gallaway, 1997).
A common concern expressed in these studies has been the potential for nearshore developments to disrupt a biologically important feature of the Beaufort Sea: a band of relatively warm, brackish water (5-10[degrees]C, 10-25) that lies adjacent to the shore (Craig, 1984). This nearshore band, which results from a combination of river discharge and solar heating of shallow nearshore waters during summer, constitutes the main feeding habitat of anadromous, amphidromous, and some marine and freshwater fishes that disperse along the coastline. Its importance relates, with minor exceptions, to the limited time available for feeding, from mid-June to mid-September, during which fish must accomplish most of their annual growth and accumulate food reserves before the onset of the nine-month winter (Craig, 1989).
Interest has centered on the Endicott Development, a gravel-filled breached causeway constructed during the winter of 1984-85 in the delta of the Sagavanirktok River as a platform for developing offshore petroleum reserves (Fig. 1). Oceanographic studies have shown that under persistent east winds, a condition that promotes coastwide upwelling of cold marine bottom water, the presence of the causeway induces the formation of a wake-eddy in the lee of its western tip (Gallaway et al., 1991). The circulation pattern within the eddy transports marine bottom water, if present, to the surface, where it is advected a short distance along the shoreward side of the causeway's western leg. Some investigators expressed concern that this causeway-induced transport of cold marine water would alter the nearshore band of warm, brackish water and cause the habitat to become marine. Such alteration would make the post-construction years similar to a naturally occurring harsh year by extending the period when cold, salty water was present (Ross, 1988), and could lead to domination of the nearshore fish community by euryhaline species at the expense of stenohaline fish (Ross, 1988). Conversely, other investigators hypothesized that the frequency and intensity of the wake-eddy events that induce cold marine water to reach the surface and the extent of the area affected would be insufficient to alter either the hydrographic character of the delta or the structure of the fish community (Gallaway et al., 1991).
Few pre-development hydrographic and fish data are available from the vicinity of the Endicott Causeway. In fact, the environmental impact statement (USACE, 1984) issued prior to the construction of the causeway was based on studies from several years of monitoring of the nearby West Dock Causeway (Fig. 1) and on a single summer's data (1982) from four stations located across the Sagavanirktok Delta (Griffiths et al., 1983). Since construction of the Endicott Causeway, fish and oceanographic monitoring studies have been conducted at the four original 1982 baseline stations, providing a large database on species abundance and water temperature and salinity in the Sagavanirktok Delta (Cannon et al., 1987; Glass et al., 1990; Reub et al., 1991; Griffiths et al., 1995).
The Endicott monitoring studies have concentrated on four fish species of importance to Native subsistence and/or sport fisheries: arctic cisco (Coregonus autumnalis), broad whitefish (C. nasus), Dolly Varden (Salvelinus malma), and least cisco (C. sardinella). The abundance levels, distribution patterns, and growth rates of these four species in the Prudhoe Bay and Endicott Causeway region have been described in detail in numerous publications (Craig and Haldorson, 1981; Gallaway et al., 1983, 1989, 1997; Craig, 1984, 1989; Craig et al., 1985; Fechhelm and Fissel, 1988; Fechhelm et al., 1989, 1991, 1992, 1993, 1994, 1995, 1997; Moulton, 1989; Fechhelm and Griffiths, 1990; English, 1991; Moulton et al., 1991; Schmidt et al., 1991; Griffiths et al., 1992; Bryan and Fechhelm, 1997; Colonell and Gallaway, 1997). These studies collectively indicate that the variations in abundance levels of these anadromous and amphidromous species observed in the Prudhoe Bay region from 1985 to 1993 are the result of coastal winds (for small arctic and least cisco), a possible density-dependent mechanism (for broad whitefish), and summer feeding dispersals (for large arctic and least cisco and Dolly Varden).
Little information, however, has been published on the so-called "minor species," those not targeted by subsistence or sport fishers. This group includes the anadromous rainbow smelt (Osmerus mordax); freshwater species, arctic grayling (Thymallus arcticus) and round whitefish (Prosopium cylindraceum); and marine species, arctic cod (Boreogadus saida), saffron cod (Eleginus gracilis), fourhorn sculpin (Myoxocephalus quadricornis), and arctic flounder (Pleuronectes glacialis). Ross (1988) suggested that these species would be most affected by causeway-induced changes in temperature and salinity levels that shift local hydrographic conditions to a more marine state. Hypothetically, marine species would become more abundant, and less salinity tolerant or freshwater species would become less abundant.
We first analyzed the temperature and salinity data from 1985 to 1993 to determine if the hydrographic environment in the post-causeway years has shifted to a more marine state. We then analyzed the patterns of relative abundance of the various fish species to test the hypothesis that there have not been any changes in abundance directly related to causeway effects. We discuss the results of these analyses in terms of causeway-induced factors versus naturally occurring events.
Fish were collected at four fyke net stations in the Sagavanirktok Delta (Fig. 1) during the open water seasons from 1985 to 1993. Nets were in place continuously over 24 h and were emptied once a day, weather permitting. The four major fish species were placed in holding pens, anesthetized in a dilute solution of tricaine (MS-222), identified, measured (fork length, mm) and released after they had recovered from the effects of the anesthetic. The minor species were counted and released. Between 1985 and 1990, YSI[R] salinity/temperature meters were used to measure daily surface temperature ([degrees]C) and salinity () at each fyke net station. From 1991 to 1993, similar measurements were taken using Hydrolab Surveyor III conductivity/temperature/depth meters. Since most of the delta is quite shallow (less than 2 m deep) and hydrographic stratification is minimal (Fechhelm et al., 1991), we considered surface hydrographic measurements to be representative of the water column at each sampling site.
Numerous sites have been sampled over the years, but only the original four 1982 stations depicted in Figure 1 have been sampled in every year, except 1987, for which there are no data (Cannon et al., 1987; Glass et al., 1990; Reub et al., 1991; Griffiths et al., 1995). For consistency among years, we used only the common sampling dates 7 July to 31 August. This eight-week period covers most of the ice-free feeding season of the fish.
The basic unit in our analyses was catch per unit effort (CPUE), standardized as the mean number of fish collected at a given net per 24 h (i.e., fish24 h-1). We assumed, on the basis of mark-recapture and tagging data, that CPUE was an index to abundance, i.e., that CPUE changed proportionally with change in abundance. Gallaway et al. (1997) showed that fyke-net CPUE was an index to abundance for small, broad whitefish in the Sagavanirktok Delta by correlating mark-recapture population estimates to net CPUE. In addition, tagging studies in Prudhoe Bay between 1981 and 1993 showed that fish did not re-enter nets immediately after release, since only 0.06% of the large anadromous fish and 0.26% of the small anadromous fish were recaptured the following day in the same net from which they had originally been released.
To avoid the problems associated with "pseudoreplication," we used the seasonal mean CPUE for each net: individual daily fyke-net CPUEs and hydrographic measurements are not truly independent samples, since they are usually strongly correlated with the previous day's and the next day's values (Stewart-Oaten et al., 1986). Our method provided four estimates of abundance (from four nets) for each year and species. The daily catch data were loge-transformed (loge [CPUE + 1]) before calculation of seasonal means. Non-transformed seasonal mean surface temperatures and salinities were used for the four baseline stations.
Using the multiple general linear hypothesis (MGLH) provided by SYSTAT (Wilkinson, 1987), we conducted multiple regression and residual analyses to determine if post-causeway changes in the temperature and salinity levels were related to year or to the other hydrographic variable or to their interaction. Similarly, we looked for significant models that could explain the observed changes in fish abundance using post-causeway year, temperature and salinity levels, and their interactions. We used the Lillifores Test to determine if the CPUE data were normally distributed. The temperature and salinity data were normally distributed, as were the transformed CPUE data for arctic cod, fourhorn sculpin, and round whitefish. However, similar data for arctic flounder, arctic grayling, rainbow smelt, and saffron cod were not normally distributed.
In cases where the data were normally distributed, we developed the "best-fit" models based on the criterion of maximized regression F-ratio (i.e., minimized regression p-values) using the all-possible-regression model. For the temperature and salinity data, we ran separate regression models, using temperature or salinity as the independent variable, against year and the remaining variable (temperature or salinity) as the dependent variables, along with the interaction term. We then removed nonsignificant interaction terms and independent variables to arrive at the "best-fit" model. Separate multiple regression models were developed for each fish species. In all cases, the independent factor CPUE was regressed against the dependent factors (year, temperature, salinity) and all possible interaction terms. We then sequentially removed the nonsignificant interaction terms and independent variables to produce the "best-fit" model.
Non-normal distributions in the transformed CPUE data for arctic flounder, arctic grayling, rainbow smelt, and saffron cod precluded parametric statistical analysis; nonparametric methods were not used, because they are inappropriate for analysis of interaction terms. We inspected these data visually, described the abundance levels, and made some generalizations about the temperature-salinity-abundance relationships in the post-causeway environment.
The "best-fit" model for explaining temperature was a simple regression with year after excluding all other variables and interactions (Table 1; Fig. 2). However, the model was not significant and explained only 6% of the variation. The best model for salinity contained year, temperature, and their interaction. This model was significant but accounted for only 17% of the observed variation (Table 1, Fig. 2). Although there was a significant year effect, the negative coefficient implies a slight decline in delta salinities since the causeway's construction. This trend was in the opposite direction to that hypothesized, namely, that the nearshore waters would become more marine as a result of the causeway's presence.
Mean temperatures varied over the nine years of record, but there was no obvious trend toward a colder, more saline condition. Two years that differed substantially from the other years (1989, with high temperatures, and 1991, with low temperatures), were both post-construction years. The high 1989 temperatures resulted from extensive periods of solar radiation that warmed the shallow nearshore waters of the Alaskan Beaufort Sea. The low 1991 temperatures were due to extensive ice cover that remained in the nearshore region for much of that summer. Both cases represented coast-wide events that affected temperatures along the Alaskan Beaufort Sea coast as far east as Barter Island, approximately 200 km away (Underwood et al., 1995).
Mean yearly salinities also varied over the same post-construction period but did not show a sequential trend toward higher salinity levels. Both high (1993) and low (1991) salinity years were recorded in post-construction years. The high 1993 salinities resulted from numerous wind-generated coastal upwelling events that occurred in July and early August of that year (LGL, Alaska, unpubl. data). The low 1991 salinities were due to the extensive ice cover that persisted in the nearshore region and prevented wind-mixing of the more marine offshore waters with the brackish nearshore waters (LGL, Alaska, unpubl. data). The low 1991 salinity pattern was also observed near Barter Island and also resulted from extensive ice cover in that area (Underwood et al., 1995). There were no comparable 1993 data from Barter Island.
Arctic cod abundance varied widely among years in the post-causeway environment (Fig. 2). There were, however, no significant regression models for arctic cod abundance and year, temperature, salinity, or their interactions (Table 1), suggesting that the changes were probably independent of year and hydrographic conditions.
The "best-fit" model for fourhorn sculpin was a simple regression on salinity, after excluding all other variables and their interactions (Table 1; Fig. 2). A plot of CPUE vs. salinity showed that the significant relationship was primarily due to the four lowest abundance levels, which occurred when salinities were less than 4, and the highest, which occurred when salinities were greater than 12 (Fig. 3).
The "best-fit" model for round whitefish was based on year, temperature, and their interaction; however, the model was not significant and accounted for only 15% of the observed variation (Table 1; Fig. 4). The significant interaction term year*temperature within the model confounded the analysis of these variables. In the end, we were left with significant factors within a nonsignificant model (Table 1). Biologically, this suggests that round whitefish abundance has not been strongly affected by temperature or salinity but that temperature may play a role in conjunction with year.
Arctic flounder and rainbow smelt both exhibited dramatic stepwise increases in abundance levels approximately five years after construction of the causeway, and these high levels persisted through 1993 (Fig. 4). No corresponding
[Part 1 of 2] TABLE 1. Results of multiple regression analyses for mean seasonal temperature and salinity and year, and for mean seasonal abundance as CPUE (loge [CPUE+1]) for selected species. Listed models are "best-fit" mo dels based on the criterion of maximized regression F-ratio (minimized P-v alue). Variable or Species Source Coefficient SS df MSE Year 0.146 Temperature Regression 4.714 1 4.7 10 Residual 47.265 30 1.5 75 Salinity Year -5.034 1 Temperature -3.437 1 Year*Temperature 0.719 1 Regression 76.172 3 25.3 91 Residual 227.367 28 8.1 20 Arctic cod Year -0.312 1 Salinity -0.214 1 Year*Salinity 0.042 1 Regression 3.856 3 1.2 85 Residual 37.607 28 1.3 43 Fourhorn sculpin Salinity 0.126 1 Regression 4.844 1 4.8 44 Residual 5.554 30 0.0 19 Round whitefish Year 0.677 1 Temperature 0.508 1 Year*Temperature -0.089 1 Regression 1.734 3 0.5 78 Residual 5.761 28 0.2 06 [Part 2 of 2] TABLE 1. Results of multiple regression analyses for mean seasonal temperature and salinity and year, and for mean seasonal abundance as CPUE (loge [CPUE+1]) for selected species. Listed models are "best-fit" mo dels based on the criterion of maximized regression F-ratio (minimized P-v alue). Variable or Species F-ratio P Adjusted r2 Temperature 2.992 0.094 0.060 Salinity 0.009(1) 0.015(1) 0.007(1) 3.127 0.042(1) 0.171 Arctic cod 0.138 0.229 0.125 0.957 0.427 0.000 Fourhorn sculpin <0.001(1) 26.164 <0.001(1) 0.448 Round whitefish 0.025(1) 0.023(1) 0.033(1) 2.809 0.058 0.149 (1) Indicates significant difference at the 0.05 level.
shifts in temperature and salinity levels occurred over the same time period (see Fig. 2). The CPUE for saffron cod has shown yearly oscillations during the study period, which probably account for the non-normal CPUE data distribution for this species (Fig. 5). As with arctic flounder and rainbow smelt, no corresponding changes in the temperature and salinity levels occurred over the same time period. These results suggest that the observed changes in abundance levels for these species were independent of hydrography.
The non-normal distribution of the arctic grayling abundance data reflects the generally low levels between 1985 and 1989, followed by two years of high catches and a return to lower levels in 1992 and 1993 (Fig. 5). The high catches in 1990 appeared unrelated to temperature and salinity, since during this year neither was substantially different than in the years with low catches. However, in 1991, when catches were the highest, salinities were the lowest recorded (see above).
Wind is the primary factor governing open-water nearshore currents and temperature and salinity patterns along the Beaufort sea coast (Hufford and Bowman, 1974; Kozo, 1984; Niedoroda and Colonell, 1990; Gallaway et al., 1991; Hachmeister et al., 1991). Wind direction is primarily from the northeast and secondarily from the northwest. The prevailing east-west direction of the winds, along with river discharge and solar warming, creates a nearshore band of low-salinity or brackish water that is, on average, markedly warmer than water further offshore (Craig, 1984; Gallaway et al., 1991). The construction of the Endicott Causeway in the winter of 1984-85 caused concern that the new structure would alter the hydrographic characteristics of the nearshore band of water, in turn affecting its suitability for fish, and as a result changing the composition of the fish community.
The "causeway effect" is the result of a combination of wind-induced, coast-wide events and the local bathymetry around the causeway (Gallaway et al., 1991). Under persistent and strong east winds, coastal surface Beaufort Sea waters are transported offshore and are replaced by a return flow of cold, salty, bottom water (temperature < 2.0[degrees]C; salinity >= 25). These periodic upwellings bring marine bottom water to the surface along the coast where depths are 2 to 3 m or less (Gallaway et al., 1991). In the vicinity of the causeway, these events typically reach the surface for 3-5 days four to five times per open-water season, accounting for about 30% of that season (Niedoroda and Colonell, 1990; Gallaway et al., 1991).
Because its orientation is perpendicular to the currents generated by east winds, the causeway acts as a barrier to water movement, causing an eddy to form in the wake of its western tip (Gallaway et al., 1991). The circulation within this wake-eddy transports bottom water to the surface, where it is advected a short distance along the inside of the causeway. During upwelling events, the water column in the wake-eddy is nonstratified and consists entirely of the cold, salty, upwelled water that reached the surface naturally. The impact of the Endicott Causeway occurs only when the water mass in the area of the wake-eddy exhibits strong vertical stratification, an event caused by the warmer, fresher water from the Sagavanirktok River overlying the cold, marine bottom layer (Niedoroda and Colonell, 1990). In these cases, the wakeeddy transports the cold marine bottom water to the surface, where it is advected along the inside of the causeway. This situation occurs in about 4% of the upwelling events, and this low frequency provides some index of the magnitude of the effects from the Endicott Causeway (Gallaway et al., 1991). Under west winds, surface waters are transported onshore; cold marine bottom water is downwelled away from shore and is not available for transport to the surface (Gallaway et al., 1991; Hachmeister et al., 1991).
On the basis of the 1985-87 Endicott oceanographic studies, it was hypothesized that this causeway-induced impact would displace the nearshore band of warm, brackish water in this area, causing the habitat to become marine (Ross, 1988). Thus, each post-construction year would be hydrographically similar to a naturally occurring harsh year, with an extended period when cold, salty water was present. Presumably, stenohaline fish would be excluded from this important feeding habitat and replaced by euryhaline species, thus altering the species community in the area. However, continued monitoring studies (from 1988 to 1993) showed that high salinity waters (> 25) rarely reached the surface, and for most of the season, high-salinity bottom water was not present in the lee of the causeway (Gallaway et al., 1991). Although the temperature and salinity patterns in the post-causeway period have varied, we found no significant relationship with year for temperature, and the significant relationship for year with salinity was in the opposite direction to that hypothesized by Ross (1988). In the case of temperature, despite the lack of a statistically significant difference, there appears to be a trend towards increasing temperatures with time. This trend, if real, is also opposite to that hypothesized by Ross (1988). What is clear is that the waters of the Sagavanirktok Delta have not become colder and more saline since the causeway was built, but rather have reflected coast-wide meteorological and oceanographic events that vary naturally among years.
Although post-construction trends in hydrography were nonsignificant for temperature and in the opposite direction for salinity, selected fish species abundance levels varied, sometimes showing dramatic increases with time. Ross (1988) hypothesized that site-specific changes in temperature and salinity at a given time, rather than average regional or seasonal changes, are of most immediate importance to the fish and may result in detrimental impacts over the years. Such changes could reduce the abundance of amphidromous and freshwater species and increase that of marine and brackish water species. The 1985-93 results do not show these patterns. The following species accounts discuss the available data in light of known life histories and measured hydrographic events.
Arctic Cod Abundance
Arctic cod is a widely distributed and abundant species in northern North American waters, occurring from the Bering Sea in the west, across the Arctic and down to the Labrador Sea in the east (Craig et al., 1982; Bradstreet et al., 1986). It is a key trophic species because of its importance in the diets of marine mammals, birds, and other fish species (Craig et al., 1982; Bradstreet et al., 1986). In the Beaufort Sea, the importance of nearshore habitats for young-of-the-year and juvenile arctic cod has been the subject of debate. Some authors have suggested that the warmer, lower-salinity nearshore waters are essential for arctic cod productivity (Cannon et al., 1991); others have found young-of-the-year, juvenile, and adult arctic cod in cold marine waters from 50-150 km offshore during the open-water season (Hunter, 1979; Lowry and Frost, 1981) and at least 175 km offshore during winter (Craig et al., 1982), which suggests a preference for offshore marine waters. Craig et al. (1982) suggested that arctic cod were infrequent visitors to nearshore habitats early in the open-water season, but that late in the summer some migrate into coastal waters. Arctic cod appear able to utilize a variety of habitats, a characteristic reflected by their wide distribution.
The abundance of arctic cod increased in the first three years after construction of the Endicott Causeway (Cannon et al., 1991), leading some to think that because of causeway-induced changes in local hydrography, arctic cod were moving into the nearshore habitat around the causeway and displacing other species (Cannon et al., 1987; Glass et al., 1990; Reub et al., 1991). Although arctic cod abundance varied between 1985 and 1993, it did not increase over time (Fig. 2). Arctic cod abundance has also been highly variable at other Beaufort Sea locations (Craig et al., 1982; Underwood et al., 1995). Craig et al. (1982) reported an almost 200-fold increase in the numbers of arctic cod between 1977 and 1978 (767 vs. 139 790) in Simpson Lagoon, about 40 km west of Endicott. They also reported that this increase was not restricted to Simpson Lagoon but was observed elsewhere along the Beaufort Sea coastline. Arctic cod were also reported to be common in the Prudhoe Bay region in 1975 and 1976 (Bendock, 1977). In the Barter Island region, approximately 200 km east of Prudhoe Bay, there were significant differences in the among-year comparisons of arctic cod abundance between 1988 and 1991 (Underwood et al., 1995). These results suggested that the arctic cod abundance varies widely along the Beaufort Sea coast, and this same variability was evident in the Prudhoe Bay region both before and after construction of the causeway. Arctic cod do not appear to have exhibited a shift in habitat use as a result of causeway-induced changes in the temperature and salinity levels in the Sagavanirktok Delta area.
Fourhorn Sculpin Abundance
Fourhorn sculpin have a circumpolar distribution (Morrow, 1980). They are ubiquitous in virtually all Alaskan Beaufort Sea nearshore habitats, including the deeper waters not frequented by anadromous and amphidromous species (Craig et al., 1985; Cannon et al., 1987; Underwood et al., 1995). Fourhorn sculpin are not a valued subsistence or recreational resource and are considered a minor species despite their widespread distribution and abundance. Most studies provide only abundance-level data for this species. However, as they are a marine species, there was concern that causeway-induced changes in local hydrography could enhance conditions for fourhorn sculpin at the expense of stenohaline species (Ross, 1988).
Fourhorn sculpin abundance was significantly and positively related to salinity (Table 1). If the water in each post-construction year had tended to be more marine, then we would have expected consistently high abundance levels for this species. However, as reported, post-construction salinity patterns and fourhorn sculpin abundances have varied synchronously (Fig. 2), suggesting that causeway-induced effects do not appear to have altered the fourhorn sculpin abundance in the area.
Round Whitefish Abundance
Round whitefish is a freshwater species found throughout mainland Alaska from the panhandle north to the Arctic coast (Morrow, 1980). It occasionally ventures into nearshore coastal waters if salinities are very low (Craig, 1984). The Sagavanirktok River supports a population of round white-fish, and they are typically captured in low numbers at stations near the delta (Cannon et al., 1987; Glass et al., 1990; Reub et al., 1991; Griffiths et al., 1995). If causeway-related or natural salinity levels had increased, we would have expected the abundance of round whitefish to decrease to the advantage of the more marine species. In this study, round whitefish abundance was significantly correlated with year and temperature, but this result was confounded by a significant year*temperature interaction (Table 1). A possible explanation is that the number of round whitefish entering nearshore waters varies each year, but once they are in nearshore waters, temperature is a major factor in determining their distribution and abundance. The important fact is that this species has not been excluded from the nearshore region, although its abundance has varied with time (Fig. 4). Round whitefish abundance did not show a continuous decrease in the post-construction years and appears to be independent of any causeway-induced hydrographic changes.
Arctic Flounder Abundance
The arctic flounder is a marine species that has a near circumpolar distribution (Morrow, 1980). Typically found in shallow coastal waters during the open-water season, it commonly enters low-salinity habitats (Walters, 1955; Morrow, 1980). Between the Colville River and Prudhoe Bay, abundance levels have been low compared to other marine, amphidromous, and anadromous species (Craig and Haldorson, 1981; Moulton et al., 1986; Cannon et al., 1987; Glass et al., 1990; Reub et al., 1991). We also found arctic flounder catch rates were relatively low during the first five post-construction years.
Starting in 1990, there was a step-change increase in CPUE that continued through 1993 (Fig. 4). A similar abrupt increase in abundance also occurred in 1990 near Barter Island, Alaska (Underwood et al., 1995). The length-frequency and age data from both areas suggest that the abundance increases resulted from a single, coast-wide event that was possibly linked to a high overwinter survival rate of the strong 1989 year-class. Most of the arctic flounder in the Endicott area in 1990 and thereafter are representatives of this year-class (LGL Alaska, unpubl. data).
The dramatic coast-wide increase in arctic flounder abundance five years after construction of the Endicott Causeway suggests that the increase was not caused by localized hydrographic patterns.
Rainbow Smelt Abundance
In Alaska, rainbow smelt are distributed from the Bering Sea north into the Arctic Ocean (Morrow, 1980). Little information on this species is available for the central Alaskan Beaufort Sea. Craig and Haldorson (1981) reported that rainbow smelt in this region are slow-growing and longlived, reaching maximum sizes of 300 mm at ages 13 to 15. Earlier studies in the Prudhoe Bay area suggested that most of the rainbow smelt in the nearshore waters during the open-water season are small, immature fish (Craig et al., 1985).
The abundance patterns of rainbow smelt between 1985 and 1993 were similar to those of arctic flounder (Fig. 4). Catch rates were relatively low in the first five post-construction years, but abundance increased sharply in 1990 and remained high thereafter. This increase in abundance appears to have been a coast-wide event, as catch data from the Barter Island area also showed a trend of increasing abundance levels starting in 1990 (Underwood et al., 1995).
As with arctic flounder, rainbow smelt population increases appear to have been due to coast-wide phenomena, and were probably not correlated with causeway construction in the Prudhoe Bay region.
Saffron Cod Abundance
The Beaufort Sea is the northern limit of saffron cod distribution (Morrow, 1980; Craig and Haldorson, 1981). Although saffron cod are widespread along the Alaskan Beaufort Sea coast, two earlier studies in our study region had reported only low numbers, typically small fish 80-220 mm in length (Bendock, 1977; Craig and Haldorson, 1981). Our results showed that saffron cod abundance varied (Fig. 5), but numbers remained relatively low in all years compared to other marine species. No comparable data on the abundance of saffron cod are available for the Barter Island region. There are no obvious relationships between the abundance of this species and temperature or salinity levels (Figs. 2 and 5) in the Prudhoe Bay region.
Arctic Grayling Abundance
The arctic grayling is a freshwater species that inhabits most Alaskan North Slope rivers (Craig and McCart, 1975). In summer, they periodically enter brackish lagoons along the Beaufort Sea coast where salinities are 4 or less, presumably to feed and to migrate from one river system to another (West et al., 1992). Because the species is stenohaline, most of these forays occur early in the open-water season, when nearshore salinity levels are low.
As with round whitefish, if salinity levels had increased we would have expected grayling to be displaced from nearshore waters and replaced by more euryhaline species. During the 1985-93 monitoring program, grayling abundance levels varied with time, with low abundance from 1985 to 1989, followed by two years of increased abundance and then two more years of low abundance (Fig. 5). These fluctuations in arctic grayling abundance do not appear to be related to any causeway-induced hydrography changes.
The nine-year time series of post-construction fish species composition and hydrographic conditions suggests that there has not been a shift in the abundance of the marine or freshwater species in the Prudhoe Bay region. There was no significant relationship between temperature and year within this period, while the significant salinity model suggested that salinities had moved in the direction opposite to that hypothesized. Several fish species showed changes in abundance levels in the post-causeway construction environment. The two clear cases involve the increased abundance of arctic flounder, a marine species, and rainbow smelt, an anadromous species. Both species appear to have undergone a sharp increase in abundance around 1989-90, and catch levels have remained high since then. Age data for arctic flounder and catch data for both species from other surveys in the Beaufort Sea suggest that the increases in abundance were broad-scale events along the Alaskan Beaufort Sea coast.
Fluctuations in fish species abundance in the Prudhoe Bay region over the period 1985-93 do not appear to have led to the hypothesized, causeway-induced shift toward dominance by marine species and a displacement of the less salt-tolerant species. Stenohaline freshwater species are still present in the area. Our data suggest that the year-to-year changes in abundance levels of fish are most likely the result of either naturally occurring events in the species' life histories or changes in coast-wide meteorological conditions, rather than a response to causeway-induced changes in the local temperature and salinity patterns.
We thank Dr. Kirk Lolman, Mike Link, and Jean Erwin for critically reviewing the manuscript. We also thank Dr. James Reynolds and two anonymous reviewers who provided many useful comments that greatly improved the manuscript. Monitoring studies were funded by the Endicott Development owners; Endicott is operated by BP Exploration (Alaska) Inc. We thank BP Exploration (Alaska) Inc. and the North Slope Borough for their continuing interest in determining the effects of coastal developments in northern Alaska. The conclusions and opinions expressed in this paper do not necessarily represent those of any of the above organizations.
BENDOCK, T. 1977. Beaufort Sea estuarine fishery study. In: Environmental Assessment. Alaskan Continental Shelf. Boulder, Colorado: Bureau of Land Management, National Oceanographic and Atmospheric Administration and OCSEAP. Final Report of Principal Investigators 4:670-729.
BRADSTREET, M.S.W., FINLEY, K.J., SEKERAK, A.D., GRIFFITHS, W.B., EVANS, C.R., FABIJAN, M.F., and STALLARD, H.E. 1986. Aspects of the biology of arctic cod (Boreogadus saida) in arctic marine food chains. Canadian Technical Report of Fisheries and Aquatic Sciences No. 1491. 193 p.
BRYAN, J.D., and FECHHELM, R.G. 1997. Use of a stress index to estimate temperature and salinity stress in arctic cisco. In: Reynolds, J., ed. Fish ecology in Arctic North America. American Fisheries Society Symposium 19:262-273.
CANNON, T.C., GLASS, D.R., ADAMS, B.A., and NELSON, T. 1987. Fish distribution and abundance. In: 1985 Final Report for the Endicott Environmental Monitoring Program. Vol. 6, Chapter 1. Prepared by Envirosphere Company for the U.S. Army Corps of Engineers, Alaska District, Anchorage, Alaska.
CANNON, T.C., GLASS, D.R., and PREWITT, C.M. 1991. Habitat use patterns of juvenile arctic cod in the coastal Beaufort Sea near Prudhoe Bay, Alaska. American Fisheries Society Symposium 11:157-162.
COLONELL, J.M., and GALLAWAY, B.J. 1997. Wind-driven transport and dispersion of age-0 arctic ciscoes along the Alaskan Beaufort Coast. In: Reynolds, J., ed. Fish ecology in Arctic North America. American Fisheries Society Symposium 19:90-103.
CRAIG, P.C. 1984. Fish use of coastal waters of the Alaskan Beaufort Sea: A review. Transactions of the American Fisheries Society 113:265-282.
_____. 1989. An introduction to anadromous fishes in the Alaskan Arctic. Biological Papers of the University of Alaska 24:27-54.
CRAIG, P.C., and HALDORSON, L. 1981. Beaufort Sea barrier island-lagoon ecological process studies: Final report, Simpson Lagoon. Part 4, Fish. Research Unit 467. In: Environmental Assessment. Alaskan Continental Shelf. Boulder, Colorado: Bureau of Land Management, National Oceanographic and Atmospheric Administration and OCSEAP. Final Report of Principal Investigators 7:384-678.
CRAIG, P.C., and McCART, P. 1975. Classification of stream types in Beaufort Sea drainages between Prudhoe Bay, Alaska and the Mackenzie Delta, N.W.T. Arctic and Alpine Research 7:183-198.
CRAIG, P.C., GRIFFITHS, W., HALDORSON, L., and McELDERRY, H. 1982. Ecological studies of arctic cod (Boreogadus saida) in Beaufort Sea coastal waters, Alaska. Canadian Journal of Fisheries and Aquatic Sciences 39:395-406.
_____. 1985. Distributional patterns of fishes in an Alaskan arctic lagoon. Polar Biology 4:9-18.
ENGLISH, K.K. 1991. Effects of temperature, salinity and prey abundance on the growth of arctic cisco and broad whitefish feeding on epibenthic prey in field enclosures. American Fisheries Society Symposium 11:119-131.
FECHHELM, R.G., and FISSEL, D.B. 1988. Wind-aided recruitment of Canadian arctic cisco (Coregonus autumnalis) into Alaskan waters. Canadian Journal of Fisheries and Aquatic Science 45:906-910.
FECHHELM, R.G., and GRIFFITHS, W.B. 1990. The effect of wind on the recruitment of Canadian arctic cisco (Coregonus autumnalis) into the central Alaskan Beaufort Sea. Canadian Journal of Fisheries and Aquatic Science 47:2164-2171.
FECHHELM, R.G., BAKER, J.C., GRIFFITHS, W.B., and SCHMIDT, D.R. 1989. Localized movement patterns of least cisco (Coregonus sardinella) and arctic cisco (C. autumnalis) in the vicinity of a solid-fill causeway. Biological Papers of the University of Alaska 24:75-106.
FECHHELM, R.G., FECHHELM, J.D., HERLUGSON, C.J., and BEAUBIEN, D.K. 1991. Bioassay tests of acute temperature and salinity shock to arctic cisco. American Fisheries Society Symposium 11:109-118.
FECHHELM, R.G., DILLINGER, R.E., Jr., GALLAWAY, B.J., and GRIFFITHS, W.B. 1992. Modeling of in situ temperature and growth relationships for yearling broad whitefish in Prudhoe Bay, Alaska. Transactions of the American Fisheries Society 121:1-12.
FECHHELM, R.G., FITZGERALD, P.S., BRYAN, J.D., and GALLAWAY, B.J. 1993. Effect of salinity and temperature on the growth of yearling arctic cisco (Coregonus autumnalis) of the Alaskan Beaufort Sea. Journal Fisheries Biology 43:463-474.
FECHHELM, R.G., BRYAN, J.D., GRIFFITHS, W.B., WILSON, W.J., and GALLAWAY, B.J. 1994. Effect of coastal winds on the summer dispersal of young least cisco (Coregonus autumnalis) from the Colville River to Prudhoe Bay, Alaska: A simulation model. Canadian Journal of Fisheries and Aquatic Science 51:890-899.
FECHHELM, R.G., GRIFFITHS, W.B., WILSON, W.J., GALLAWAY, B.J., and BRYAN, J.D. 1995. Intra- and interseasonal changes in the relative condition and proximate body composition of broad whitefish from the Prudhoe Bay region of Alaska. Transactions of the American Fisheries Society 124:508-519.
FECHHELM, R.G., BRYAN, J.D., GRIFFITHS, W.B., and MARTIN, L.R. 1997. Summer growth patterns of northern Dolly Varden (Salvelinus malma) smolts from the Prudhoe Bay region of Alaska. Canadian Journal of Fisheries and Aquatic Sciences 54:1103-1110.
GALLAWAY, B.J., GRIFFITHS, W.B., CRAIG, P.C., GAZEY, W.J., and HELMERICKS, J.W. 1983. An assessment of the Colville River delta stock of arctic cisco--migrants from Canada? Biological Papers of the University of Alaska 21:4-23.
GALLAWAY, B.J., GAZEY, W.J., and MOULTON, L.L. 1989. Population trends for the arctic cisco (Coregonus autumnalis) in the Colville River of Alaska as reflected by the commercial fishery. Biological Papers of the University of Alaska 24:153-165.
GALLAWAY, B.J., GAZEY, W.J., COLONELL, J.M., NIEDORODA, A.W., and HERLUGSON, C.J. 1991. The Endicott Development Project-Preliminary assessment of the impacts from the first major offshore oil development in the Alaskan Arctic. American Fisheries Society Symposium 11:42-80.
GALLAWAY, B.J., FECHHELM, R.G., GRIFFITHS, W.B., and COLE, J.G. 1997. Population dynamics of broad whitefish in the Prudhoe Bay region, Alaska. In: Reynolds, J., ed. Fish ecology in Arctic North America. American Fisheries Society Symposium 19:194-207.
GLASS, D., WHITMUS, C., and PREWITT, M. 1990. Fish distribution and abundance. Final report for the 1986 Endicott Environmental Monitoring Program. Prepared by Envirosphere Company for the U.S. Army Corps of Engineers, Alaska District, Anchorage, Alaska. Part IV, Vol. 5, Chapter 1.
GRIFFITHS, W.B., SCHMIDT, D.R., FECHHELM, R.G., GALLAWAY, B.J., DILLINGER, R.E., Jr., GAZEY, W., NEILL, W.H., and BAKER, J.S. 1983. Fish ecology. In: Gallaway, B.J., and Britch, R., eds. Vol. 3. Environmental Summer Studies (1982) for the Endicott Development. Report by LGL Alaska Research Associates, Inc. and Northern Technical Services for Sohio Alaska Petroleum Co., Anchorage, Alaska. 342 p.
GRIFFITHS, W.B., GALLAWAY, B.J., GAZEY, W.J., and DILLINGER, R.E., Jr. 1992. Growth and condition of arctic cisco and broad whitefish as indicators of causeway-induced effects in the Prudhoe Bay region, Alaska. Transactions of the American Fisheries Society 121:557-577.
GRIFFITHS, W.B., FECHHELM, R.G., MARTIN, L.R., WILSON, W.J., and COLONELL, J.M. 1995. The 1994 Endicott Development Fish Monitoring Program. Vol. I: Fish and hydrography data report. Unpubl. report for BP Exploration (Alaska) Inc. and North Slope Borough. 182 p. + appendices. Available from LGL Alaska, 4175 Tudor Centre Drive, Suite 101, Anchorage, Alaska 99508, U.S.A.
HACHMEISTER, L.E., GLASS, D.R., and CANNON, T.C. 1991. Effects of solid-fill gravel causeways on the coastal Beaufort Sea. American Fisheries Society Symposium 11:81-96.
HUFFORD, G.L., and BOWMAN, R.D. 1974. Airborne temperature survey of Harrison Bay. Arctic 27:69-70.
HUNTER, J.G. 1979. Abundance and distribution of arctic cod, Boreogadus saida, in the southeastern Beaufort Sea. Canadian Atlantic Fisheries Scientific Committee Research Document No. 79/39. 13 p.
KOZO, T.L. 1984. Mesoscale wind phenomena along the Alaskan Beaufort Sea coast. In: Barnes, P.W., Schell, D.M., and Reimnitz, E., eds. The Alaskan Beaufort Sea ecosystems and environments. New York: Academic Press. 23-45.
LOWRY, L.L., and FROST, K.J. 1981. Distribution, growth and foods of arctic cod (Boreogadus saida) in the Bering, Chukchi and Beaufort Seas. Canadian Field-Naturalist 95:186-191.
MORROW, J.E. 1980. The freshwater fishes of Alaska. Anchorage, Alaska: Alaska Northwest Publishing Company. 284 p.
MOULTON, L.L. 1989. Recruitment of arctic cisco (Coregonus autumnalis) into the Colville Delta, Alaska, in 1985. Biological Papers of the University of Alaska 24:107-111.
MOULTON, L., GALLAWAY, B., FAWCETT, M., GRIFFITHS, W., CRITCHLOW, K., FECHHELM, R., SCHMIDT, D., and BAKER, J. 1986. 1984 Central Beaufort Sea fish study. Waterflood Monitoring Program Fish Study Report prepared for Envirosphere Co., Anchorage, Alaska. 300 p.
MOULTON, L.L., FIELD, L.J., and KOVALSKY, R. 1991. Predictability in the catch of arctic cisco in the Colville River, Alaska. American Fisheries Society Symposium 11:146-156.
NIEDORODA, A.W., and COLONELL, J.M. 1990. Beaufort Sea causeways and coastal ocean dynamics. In: Chakrabartl, S.K., Magda, H., Aage, C., and Nielsen, F.G., eds. American Society of Mechanical Engineers Proceeding of the Ninth International Conference of Offshore Mechanical and Arctic Engineering. Book No. 10296B. 509-516.
REUB, G.S., DURST, J.D., and GLASS, D. 1991. Fish distribution and abundance. Final report for the 1987 Endicott Environmental Monitoring Program. Prepared by Envirosphere Company for the U.S. Army Corps of Engineers, Alaska District, Anchorage, Alaska. Part IV, Vol. 6, Chapter 1.
ROSS, B.D. 1988. Causeways in the Alaskan Beaufort Sea. U.S. Environmental Protection Agency Report 910/0-88-218, Region 10, Seattle, Washington.
SCHMIDT, D.R., GRIFFITHS, W.B., BEAUBIEN, D.K., and HERLUGSON, C.J. 1991. Movement of young-of-the-year arctic ciscoes across the Beaufort Sea coast, 1985-1988. American Fisheries Society Symposium 11:132-144.
STEWART-OATEN, A., MURDOCK, W.M., and PARKER, K.R. 1986. Environmental impact assessment: "Pseudoreplication" in time? Ecology 67:929-940.
UNDERWOOD, T.J., GORDON, J.A., MILLARD, M.J., THORPE, L.A., and OSBORNE, B.M. 1995. Characteristics of selected fish populations of Arctic National Wildlife Range coastal waters. Final Report, 1988-1991. Fairbanks Fishery Resource Office, Alaska Fisheries Technical Report Number 28. Fairbanks, Alaska: U.S. Fish and Wildlife Service. 590 p.
(USACE) UNITED STATES ARMY CORPS OF ENGINEERS, ALASKA DISTRICT. 1984. Endicott Development Project Final Environmental Impact Statement. Vol. II: Technical Discussion. Available from U.S. Army Corps of Engineers, Pouch 898, Anchorage, Alaska 99506, U.S.A.
WALTERS, V. 1955. Fishes of western arctic America and eastern arctic Siberia. Bulletin of the American Museum of Natural History 106:225-368.
WEST, R.L., SMITH, M.W., BARBER, W.E., REYNOLDS, J.B., and HOP, H. 1992. Autumn migration and overwintering of arctic grayling in coastal streams of the Arctic National Wildlife Refuge, Alaska. Transactions of the American Fisheries Society 121:709-715.
WILKINSON, L. 1987. SYSTAT: The system for statistics. Evanston, Illinois: SYSTAT, Inc.
WILSON, W.J., and GALLAWAY, B.J. 1997. Synthesis in applied fish ecology: Twenty years of studies on effects of causeway development on fish population in the Prudhoe Bay region, Alaska. In: Reynolds, J., ed. Fish ecology in Arctic North America. American Fisheries Society Symposium 19:326-339.
|Printer friendly Cite/link Email Feedback|
|Author:||William B. Griffiths, and others|
|Date:||Jun 1, 1998|
|Previous Article:||Caribou distribution during the post-calving period in relation to infrastructure in the Prudhoe Bay Oil Field, Alaska.|
|Next Article:||Paleoethological reconstruction and taphonomy of Equus lambei from the Bluefish Caves, Yukon Territory, Canada.|